Energy Evaluation for Cloud RAN Employing TDM-PON as Front-Haul Based on a New Network Traffic Modeling

Abstract

Cloud radio access network (C-RAN) is a novel network architecture to meet the ever-increasing user demands. Specially, the operating expense of C-RAN, including the energy consumption, is an issue of common concern. Former studies in energy consumption evaluation of C-RAN only considered the case of one virtual server and one remote radio head, which ignored the optical front-haul and cannot provide an overall analysis for the whole access network. In this paper, we analyze the energy consumption of the whole C-RAN architecture, employing Ethernet-based time division multiplexing passive optical network (TDM-PON) (i.e. 10G EPON) as the optical front-haul. Obviously, it becomes complex to analyze the energy problem for the whole network after bringing in the front-haul part. Based on queueing theory, this paper provides a new network traffic modeling for energy analysis in C-RAN. And, we propose a strategy to make the system work in the most energy-efficient state considering the quality of service. In addition, the comparison between C-RAN and 10G-EPON-LTE architecture has also been conducted. We analyze the power-delay tradeoff, the power changes along with request arrival rate, and the daily power distribution in the two networks. From the numerical results, C-RAN architecture can save more than 60% power in a typically low request arrival rate, and reduce about 40% energy for a whole day, considering the daily user behavior, which demonstrates its large potential in the next generation network.